System Identification of a Quarter Car Model

How to run the code?

The exploration considers various parameters to be swept to understand estimation method performance. The parameters include:

1. The input road relative velocity
2. The sampling time of the estimation data
3. Simulation & Prediction Focus setting of the model The script quarterCarIdParameterSweep runs the entire parameter sweep by calling the generateResponse and identifySystem functions. The variables velRange , sampTimeRange and numExpPerInput define the range of experiments to be conducted. Currently, the code is set to perform estimation for velRange of 1 kmph to 20 kmph under sampTimeRange of 0.01s to 0.06s and running 20 experiments for each setting.

Note that the code uses parfor and parsim functionality to speed up processing. Therefore, it is necessary to have MATLAB Parallel Computing Toolbox installed on your machine to run the code. On an octacore 9th gen Intel i9 workstation, this small sweep took almost 5 hours, so consider your machine bandwidth before running the whole range.

The results from the parameter sweep are stored in a separate qcmVariables.mat file for post-processing. All the plots are generated in the Plots live script.